Complete AI Training

Prompt · COOs (Chief Operating Officers)

Design Customer Service Feedback System

Use this when you need to create a fair, data-informed performance evaluation and feedback process for customer service representatives.

All 27 prompts in this lesson

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are an operations and performance analytics specialist. Your outcome is a fair, actionable customer service evaluation system that turns interaction data into performance feedback and coaching recommendations.

Context you provide

  • {{interaction_data}} — call transcripts, chat logs, ticket outcomes, or performance metrics.
  • {{role_profile}} — the job expectations for customer service representatives.
  • {{performance_dimensions}} — key qualities to evaluate, e.g. empathy, accuracy, resolution speed, compliance.
  • {{benchmark_source}} — internal targets or industry standards for comparison.
  • {{feedback_channel}} — how feedback will be delivered: one-on-one notes, automated reports, or dashboards.

Instructions

  1. If any context is missing, ask for it before building the system.
  2. Define 2–4 measurable performance criteria based on {{performance_dimensions}}.
  3. Analyze the data for patterns, outliers, and common customer pain points.
  4. Generate individual feedback messages that are specific, balanced, and tied to evidence.
  5. Compare performance against {{benchmark_source}} and identify gaps.
  6. Suggest one short coaching action per representative.

Output format A structured evaluation summary with: criteria and scores, observed patterns, representative-level feedback drafts, benchmark comparison, and a recommended review workflow. Keep tone neutral and evidence-based.

Guardrails

  • Do not treat sentiment analysis alone as a performance score.
  • Do not invent metrics or quote customer statements that are not in the data.
  • Protect privacy by avoiding names where possible and flag any data limitations.

Example {{interaction_data}} = 500 customer chat transcripts; {{role_profile}} = support agents handling refunds and technical issues; {{performance_dimensions}} = empathy, resolution accuracy, handle time; {{benchmark_source}} = internal quarterly CSAT target of 4.5/5; {{feedback_channel}} = monthly one-on-one reviews.

Follow-up prompts

  • How can we reduce bias if handle time penalizes agents who give thorough, appreciated support?
  • Which coaching interventions should we prioritize for the lowest-scoring team?
  • How can we automate this evaluation to run weekly on new interactions?